I hit a wall. Hard. It was 2022, and I was running my second company, RemoteTeam. We were growing fast, but I was drowning. My calendar was a chaotic mess of back-to-back meetings. My inbox was a relentless firehose of demands. I was working 16-hour days, fueled by caffeine and the gnawing fear that if I stopped, even for a second, everything would fall apart. I was the classic Silicon Valley founder: running on empty, pretending I had it all under control. I didn’t.
Then we got acquired by Gusto. You’d think that would be the moment to breathe. To relax. But the habits of burnout are hard to break. I just found new things to fill the void. I started angel investing, advising startups, and writing a book. Before I knew it, I was right back where I started: overwhelmed and on the fast track to another burnout.
That’s when I decided to go all-in on AI. Not just dabbling with a few tools, but systematically trying to build an AI-powered personal operating system. A way to offload the cognitive clutter and get back to the work that actually matters. I’ve invested in over 200 companies, including some of the biggest names in AI like Anthropic, OpenAI, Scale AI, and Hugging Face, so I had a front-row seat to the revolution. I figured it was time to start using the tools I was funding.
Some of my experiments were a complete disaster. Others were life-changing. I’m opening up my playbook to show you what worked, what didn’t, and how to build your own AI-powered personal OS. Before you spend another dollar on a fancy AI subscription, read this. I tested the top 7 tools, and the winner wasn’t what I expected at all.
1. Your AI Is Only as Good as Your Data
This sounds obvious, but it’s the single biggest mistake people make. They sign up for a shiny new AI tool, expect it to read their mind, and then get frustrated when it spits out generic garbage. The truth is, these models are like incredibly smart, but incredibly naive, interns. You have to train them. You have to give them context.
My first attempt at an AI assistant was a catastrophe. I tried to get it to draft emails for me. I fed it a few examples of my writing and thought that would be enough. The results were embarrassing. The tone was all wrong. It sounded like a corporate robot. It was so bad, I had to apologize to a few people for the weirdly formal emails they received from me.
I learned you have to be relentless about feeding your AI. I now have a system where I BCC my AI on every email I send. I’ve connected it to my personal CRM. I’ve given it access to my meeting notes, my documents, my entire digital life. It’s a bit scary, I’ll admit. But now, when I ask it to draft an email, it sounds like me. It knows my relationships. It knows my priorities. It’s not just an AI, it’s my AI.
2. Don't Automate Everything
The temptation with AI is to automate everything. To create a world where you never have to do any "work" at all. This is a trap. Some tasks are too important to automate. Some tasks are where the real thinking happens.
I tried to automate my deal flow analysis for angel investments. I built a system that would scrape pitch decks, analyze the market, and give me a "go" or "no-go" recommendation. It was a technical marvel. It was also completely useless.
The problem is that investing is about people. It’s about the founder’s vision. It’s about the team’s chemistry. These are things that an AI can’t (yet) understand. My fancy system was optimizing for the wrong things. It was telling me to invest in companies that looked good on paper but lacked the human element that actually drives success.
I still use AI to help with my investments. It’s great for summarizing research, for tracking companies, for managing my portfolio. But I’ll never automate the final decision. That’s my job.
3. The Best AI Tools Are Invisible
I’ve tested dozens of AI tools. The ones that have stuck are the ones I barely notice. They’re the ones that are seamlessly integrated into my existing workflows. They don’t require me to open a new app or learn a new interface. They just work.
One of my favorites is an AI-powered calendar. It’s not a separate app. It just plugs into my Google Calendar. It automatically finds meeting slots that work for everyone. It sends out invites. It reschedules conflicts. It’s saved me countless hours of back-and-forth emails. I don’t even think about it anymore. It’s just part of how I schedule my life.
This is the future of AI. Not a bunch of separate, clunky apps, but a layer of intelligence that’s woven into the fabric of everything we do. The tools that win will be the ones that disappear.
4. Free Is Often Better Than Paid
There’s a gold rush in AI right now. Everyone is launching a new tool, a new subscription, a new way to take your money. But some of the most powerful AI tools are the ones that are already built into the products you use every day. And many of them are free.
I was paying for a fancy AI writing assistant for a while. It was expensive, and honestly, not that great. Then I discovered that the built-in AI features in Google Docs and Notion were just as good, if not better. They have access to all my documents, so the suggestions are much more relevant.
Before you shell out for a new AI subscription, take a look at the tools you’re already using. You might be surprised at what you find. The winner of my 7-tool showdown wasn’t a flashy new startup. It was the humble, and free, AI features in the tools I already used every day.
5. AI for Email Is a Trap (Mostly)
This one might be controversial. Everyone is excited about AI for email. The promise of an AI that can write all your emails for you is seductive. But I think it’s a trap. Email is about relationships. It’s about communication. When you automate it, you lose the human connection.
I’m not saying AI has no place in email. It’s great for summarizing long threads. It’s great for finding that one attachment you can’t seem to locate. But I’m very wary of using it to write my emails for me. I’ve seen too many people fall into the trap of sending generic, soulless emails that do more harm than good.
My rule is this: if the email is important, write it yourself. If it’s a quick, transactional email, then maybe let the AI help. But never let the AI be the primary author of your most important communications.
6. Meeting Notes Are the Gateway Drug
If you’re going to start with one thing, start with AI for meeting notes. It’s the highest leverage, lowest risk way to get started with AI-powered productivity. Meetings are a huge time suck in most organizations. And most of the time, the notes are terrible or non-existent.
I use an AI tool that joins all my Zoom calls. It records them, transcribes them, and then creates a summary with action items. It’s been a game-changer. I can now be fully present in meetings, knowing that the AI is taking care of the notes. I can easily search back through my meeting history to find that one thing that someone said that I can’t quite remember.
This is the gateway drug because once you see how powerful AI can be for meetings, you’ll start to see the possibilities everywhere else. You’ll start to think, "If it can do this for my meetings, what can it do for my email? My documents? My entire workflow?"
7. You Need a Chief of AI Staff
You can’t just have a bunch of disconnected AI tools. You need a system. You need a central brain that connects all the dots. I call this my "Chief of AI Staff." It’s not a single tool, but a combination of tools and workflows that work together to create a cohesive system.
My Chief of AI Staff is built around a central database of all my information. My emails, my meetings, my documents, my contacts. Everything. Then I have a set of AI agents that can access this database and perform tasks on my behalf. One agent is in charge of my calendar. Another is in charge of my email. Another is in charge of my research.
This is the holy grail. A single, unified system that knows everything about you and can help you with anything. We’re not quite there yet, but we’re getting close. And the people who figure this out first will have a massive advantage.
8. The Goal Isn't to Do More, It's to Think Better
This is the most important lesson of all. The goal of all this automation isn’t just to cram more tasks into your day. It’s not about becoming a productivity machine. It’s about creating space for deep work. For creative thinking. For the things that only you can do.
I used to think that the goal was to get to "inbox zero." Now I realize that’s a fool’s errand. The goal is to spend as little time in my inbox as possible. The goal is to have more time to think, to write, to talk to interesting people. The goal is to be less busy, not more productive.
AI is a powerful tool. But it’s just a tool. It’s up to us to decide how we use it. We can use it to become even more addicted to the cult of productivity. Or we can use it to reclaim our time, our creativity, and our humanity.
I’m still on this journey. My system is far from perfect. But for the first time in a long time, I feel like I’m in control. I’m not just reacting to the world, I’m creating it. And that’s a feeling that no amount of money can buy.
Frequently Asked Questions
How do I know which items apply to my situation?
Start by honestly assessing where your biggest bottleneck is right now. The items that address that specific constraint will give you the highest return on your time and energy.
Are these recommendations still relevant in 2026?
Absolutely. While specific tools and tactics change, the underlying principles remain consistent. I update my thinking regularly based on what I'm seeing in the market and across my portfolio companies.
How were these items selected?
Each item on this list comes from direct experience, either from building my own companies or from patterns I've observed across the 200+ startups I've invested in. I prioritize practical, actionable items over theoretical concepts.